Your cognitive profile, portable to any AI.

€249 one-time. Permanent catalog access. Works in Claude, ChatGPT, Gemini, Mistral, and any serious AI.

Why commercial LLMs are sycophantic

Big LLMs are trained to validate you. It's learned behavior, not malice. But it wastes your time and, when stakes are high, gives you soft advice when you need candor. Loading your cognitive profile into the external AI changes that: you calibrate it to your voice, your tolerance for dissent, and your humor — so it treats you as a specific interlocutor, not as a generic user.

Eight modes of use

Your profile serves eight distinct purposes. Three by default, five activatable on demand.

Public modes catalog →

Differentiation

Compared to "ChatGPT with custom instructions", a bare Claude Project, ChatGPT Memory (black box), or "building personality by talking to the AI" (not psychometrically validated), AfiniTwin offers:

• Psychometric validation with recognized instruments (IPIP-NEO, ECR-R, HSQ, AVI, ZTPI).
• Explicit anti-sycophancy in the directive meta-prompt.
• Multi-mode: eight differentiated uses of the same profile.
• AI Act + GDPR compliance, documented and public.
• Real portability: download, paste, it works.
• Snapshot consultable indefinitely.

Full comparison table →

Scientific validity

Five internationally validated psychometric instruments, with adaptive norms calibrated on 1,461,383 respondents. The profile you take to your external AI is the same one Mi IA already uses inside afini.ai — not a simplified version.